Stacking stars with Hugin's help (show and tell)

273 views
Skip to first unread message

Monkey

unread,
Oct 26, 2013, 9:01:17 AM10/26/13
to hugi...@googlegroups.com
Hi all,

Attached are a couple of images showing what I was able to do with Hugin's help.

single.jpg is a single 30-second fisheye exposure of the night sky, while stack.jpg is the averaged result of about 45 such images.

I converted the RAW camera files to TIFFs in Photoshop (which is quick and has good automated hot-pixel removal), then made a bitmap of points identifying about 30 of the brightest stars. I used parts of my own previously-written library to track these stars across the sequence of images, and then generated the control points to add to a .pto file of the images - rather than linking each image to every other, this only generated points from each image to image 0.

Hugin was able to optimise the images really well, then I wrote another piece of code to average the remapped TIFFs to a 32-bit .pfm, which was then opened in Photoshop and had the levels fiddled with until I got the result attached.

The attached are scaled down to 25%. The residual circular streaking you can see is down to regular noise in the camera's sensor, as I didn't use a dark frame. About midway between the bright star in the bottom half of the image and the glowing tree on the left, you might see what looks like a small diagonal streak - it's not a flare or meteor or anything like that, but five stars in Brocchi's Cluster which are surprisingly well-aligned.

In future I plan to automate a little more of this process - perhaps writing a simple GUI for directly opening RAW files and tracking stars to create a .pto file for Hugin to optimise, and then a stacker which can work on the original RAW data (which, with enough images, means you can do away with Bayer interpolation)

David
single.jpg
stack.jpg

David W. Jones

unread,
Oct 26, 2013, 9:55:34 PM10/26/13
to hugi...@googlegroups.com
On 10/26/2013 03:01 AM, Monkey wrote:

> The attached are scaled down to 25%. The residual circular streaking you
> can see is down to regular noise in the camera's sensor, as I didn't use
> a dark frame.

Definitely do a dark frame. It will not only deal with regular noise, it
will also catch any hot pixels.

--
David W. Jones
gnome...@gmail.com
wandering the landscape of god
http://dancingtreefrog.com

David Haberthür

unread,
Oct 27, 2013, 7:09:29 AM10/27/13
to hugi...@googlegroups.com
Ciao David.
Not to 'curb your enthusiasm' with developing something yourself to ge
the nice looking images: have you looked at http://starstax.net/ ?
Or do you not want to generate startrails, but generate high quality
images of a 'non-moving' sky with averaging multiple exposures?
David/Habi
> --
> A list of frequently asked questions is available at:
> http://wiki.panotools.org/Hugin_FAQ
> ---
> You received this message because you are subscribed to the Google Groups
> "hugin and other free panoramic software" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to hugin-ptx+...@googlegroups.com.
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/hugin-ptx/5e7beab5-a9a7-4d89-9ea6-0230d6ebbea2%40googlegroups.com.
> For more options, visit https://groups.google.com/groups/opt_out.

Monkey

unread,
Oct 28, 2013, 9:08:29 AM10/28/13
to hugi...@googlegroups.com
Nope, I wanted to remap the images so as not to generate star trails. There is software around that does this, but I've never found one I'm completely happy with - either they only support simple remapping (sometimes only translation, sometimes also rotation but that isn't always enough), or the interface is bizarre, or they just plain crash.
Reply all
Reply to author
Forward
0 new messages